Naive and memory B cells in T-cell-dependent and T-independent responses

(1)

in Immunopathology

9 Springer-Verlag 2001

Naive and memory B cells in T-cell-dependent

and T-independent responses

Rudolf H. Z u b l e r

Division of Hematology, Geneva University Hospital, Switzerland

Abstract. This review focuses on the properties and roles of distinct subsets among the primary and the memory B lymphocytes regarding their contribution to helper- T-cell-dependent and -independent antibody responses. The naive/memory B cell functions are explained in the context of current concepts on the basic mechanisms of humoral immunity. Differences between murine and human B cells are also discussed.

Continuous B cell production maintains primary B repertoires

New B cells are produced at a high rate from hematopoietic stem cells until ad- vanced age in humans [59]. This maintains primary B cell repertoires for the constitutive production of natural antibodies, and for the generation of T-independent and T-dependent adaptive immune responses to antigens to which the organism becomes exposed. The V(D)J DNA rearrangements which create the primary B cell antigen receptor (BCR) specificity repertoires occur in a context of non-random utilization of variable region (V) genes reflecting the evolutionary pressures to produce antibodies reacting with frequent pathogens [1].

At the immature B cell stage, where a BCR formed of ~t heavy and kappa (or lambda) light chains becomes expressed at the cell surface for the first time, self- reactivity generates signals that lead to receptor editing: replacement of unsuitable V regions by re-rearrangement of DNA. This occurs frequently; it is only if such editing fails within an allowed time frame that apoptosis occurs [ 10]. The bone marrow environment permits editing [61]. The immature B cells then undergo final maturation in the periphery in contact with all the tissues accessible to recirculating cells. Strong self-reactivity causes apoptosis, weaker reactivity causes different degrees of inactivation called anergy. Anergic cells still can be recruited into immune responses by appropriate activation signals. A low level of self-reactivity leads to positive selection of B 1 B cells producing the natural antibodies [28]. Differentiation into this B sublineage thus seems to be instructed by BCR signals. However, all naive B cells must express a BCR to survive.

Correspondence to: Rudolf H. Zubler, Division of Hematology, Geneva University Hospital, 12ll Geneva 14, Switzerland; e-mail: rudolf.zubler@hcuge.ch

(2)

The set point of the BCR molecular signalosome [20], which determines these outcomes according to spatial density and affinity of self-determinants, as well as to B maturation stage, is influenced by signals from a variety of other receptors that modulate the BCR signal, such as CD5, CD22, FcyR2b, PD-1 receptor, etc. [18, 54]. Reduced capacity for anergy and apoptosis causes increased self-reactivity, usually together with expansion of B 1 B cells. Coexpression of IgD together with IgM defines mature B cells. IgD seems to exert subtle effects on B cell activation/tolerance, so subtle that 6 knockout is irrelevant for the life of a laboratory mouse.

Symbiosis of B cells with their m i c r o e n v i r o n m e n t s

Distinct 13 sublineages emigrating from the bone marrow express different chemokine receptors and homing molecules which direct them according to combinatorial codes to different microenvironments [ 13]. B 1 B cells migrate mainly into peritoneal and pleural tissues/cavities. Of the other ("conventional" or B2) B cells, some home to the primary B follicles of either the Peyer's patches of gut-associated lymphoid tissue (GALT) or other mucosa-associated lymphoid tissue (MALT), and/or the spleen or lymph nodes. Other B2 cells home to the splenic marginal zone and microenvironments corresponding functionally to the splenic maNinal zone between follicles in human lymph nodes and Peyer's patches [48]. The marginal zone is a macro- phage- and dendritic cell-rich microenvironment which favors T-independent B responses. It is located at the white-red pulp junction, close to the sinuses where the circulating antigens arrive. The tyrosine kinase Pyk-2 and the transcription factor Aiolus must be expressed in murine B cells for marginal zone homing [26]. In a symbiotic relationship the B cells actively support the development and maintenance of the microenvironments on which they depend. B cells are required for the development of Peyer's patches including M cells [22]. B cells produce lymphotoxin ct213 and tumor necrosis factor ct (TNFct) for the follicular stroma cells [3]. Only a minor- ity of the newly generated B cells become relatively tong-lived cells that recirculate between blood and their borne environments. All primary B cells must express an antigen receptor for survival and for homeostatic expansion to predetermined cell levels. This is also the case for naive T cells. The TCR-proximal tyrosine kinase p561ck is necessary for homeostatic proliferation, but not for survival [64].

Pre-B cells in the bone marrow express the Ia heavy chain; the splicing required for ~) expression - which experimentally can replace )a for B maturation - is normally suppressed. Mice unable to express ~ or (5 were found to generate IgA B cells quite abundantly as well as intestinal IgA antibodies [41]. In these mice, bone-marrow- derived cells mature to pre-B and further stages in the GALT, but not in the bone marrow. Conventional B cells capable of generating T-independent adaptive [gA responses as well as lgA-producing B I cells are apparently generated in the GALT. Possibly this pathway may also be important in normal mice and humans.

Natural antibodies

The natural antibodies, which are mainly IgM and, in the MALT, to a large extent IgA, as found in mice, are important as first-line defense against microorganisms [18, 41, 50]. T-independent as well as ,//~)-T-cell-dependent mechanisms participate

(3)

in the differentiation of the B I into plasma cells [78]. B1 B cells and some ,/16- T cells can be considered as part of natural immunity. The B 1 cells may be positively selected not only by self-antigens but also by determinants of the constitutive com- mensal bacterial flora in the gut. B 1 cells frequently express CD5. This inhibitory receptor [7] is also found on anergic B cells [30] and on B cells recently activated by T cells in vitro [24]. This means that the self-reactivity of the B1 cells is constantly controlled, and that CD5 is not a specific marker for B lcells. The natural antibodies generally show broad cross-reactivity with soluble and cellular self-antigens (low- affinity polyreactive antibodies); they neutralize some pathogens quite efficiently [50]. Most likely, B 1 cells can also be recruited into adaptive immune responses.

Antigen recognition in immunological synapses

Chronically persisting antigens are tolerogenic, i.e. as if they were all self-determinants. Chronic BCR ligation leads to diminished calcium signaling. This is sufficient for activation of the transcription factor NFATc, but not for that of NFI<B, the key transcription factor for lymphocyte activation. Because the stimulatory effects of NFATc mainly occur by synergism with NFKB, the inhibitory effects of this multi- functional transcription factor predominate [23]. Adaptive immune responses thus occur in response to an acute increase of antigen. The primary responses take place only in the secondary lymphoid tissues [83]. T-dependent and probably most T-independent antigens (at least in humans) are recognized by the B cells on professional antigen-presenting cells (APC), such as migrating dendritic cells (DC) or resident follicular dendritic cells (FDC), by the formation of an immunological synapse.

A synapse is formed by large membrane areas of polarized interacting cells [5, 16, 46]. The antigen receptors and the many other receptors for positive and negative signals are spatially organized, together with their signaling modules, in complex concentric configurations that change over time: first one finds the adhesion molecules and later on the antigen/antigen receptors in the central supramolecular activation cluster (c-SMAC). This requires dynamic cytoskeletal alterations and adaptors for the linkage of the signaling components. WASR the protein deficient in Wiskott- Aldrich syndrome, is such an adapter [16]. T-independent B cell responses to the polysaccharides of encapsulated bacteria are particularly affected by this condition.

One aspect of the synapses then is that positive and negative signals are organized. Phosphatase CD45, which is required for activation of antigen-receptor- associated tyrosine kinases, is at one time moved to the periphery of the T-cell-APC synapse, but later on retransported into the center by means of endosome traffic [16]. The list of cellular receptors that inhibit B cell activation is rapidly growing [18, 54]. B cell activation signals also favor expression of or signaling from inhibitory receptors [23]. One can assume that ligands on various tissues suppress B cell activation, in analogy to the protective function of CD47 on red blood cells against phagocytosis by macrophages [51]. In particular, synapsing is required for the proper activation of naive lymphocytes.

Another aspect is antigen concentration in the c-SMAC. Rare antigens bound to the APC (e.g. via complement receptors) become concentrated. The BCR become concentrated in the glycolipid-enriched membrane domains called lipid rafts, linked to their signalosomes. B cells selectively internalize the BCR that bind antigen, for processing and presentation of peptides on MHC class II to a potential helper T cell.

(4)

Even antigen presented as a transmembrane molecule on an APC can be internalized, by membrane pinching [5].

Synapses and properties of T-independent antigens

For many microbial antigens, the DC and / or B cells are equipped with receptors of innate immunity, such as the Toll-like receptors for Escherichia coli lipopolysaccha- ride (TLR4), bacterial CpG DNA (TLR9), etc. [53]. T-independent type 1 (TI-1) antigens react with such receptors. Human B cells express some toll-like receptors (TLR) [49]; TLR2 (a receptor for peptidoglycans form Staphylococcus aureus) is found on GC B cells, but not on peripheral blood B cells [19]. However, in contrast to murine B cells, human B cells do not respond to lipopolysaccharide. Costimulato- ry molecules leading to B cell activation occur on DC and FDC, e.g. DC-8 on human FDC [37]. Thus, these cells can transfer their activation state onto the B cells.

TI-2 antigens cause extensive BCR cross-linking; they are formed by arrays of repetitive determinants in a rigid form, such as occur on many bacteria and viruses [76, 83]. In contrast to naive murine B cells, in vitro stimulation of naive human B cells by extensive BCR cross-linking [anti-Ig antibodies on beads or coated plates, or pro- tein-A-rich S. aureus Cowan I (SAC)], alone or together with many cytokines, in- cluding BAFF/BIyS [63], does not lead to Ig secretion, only some proliferation. Even CD40 ligation together with B-differentiation-inducing cytokines - in the absence of intact T cells - does not induce Ig secretion in naive B cells [31]. Memory B cells can be readily induced by SAC to proliferate and secrete Ig [44, 80]. Therefore it ap- pears that the activation of naive B cells in particular is more tightly controlled in humans than in mice, probably because organisms with a longer generation time must exercise a better control over autoimmunity. This may require synapse formation.

How B cell activation by TI-2 relates to synapse formation remains to be eluci- dated. Manifestly, some antigens induce TI-2 responses better than others, independent of the antigen close / antigen concentration in the c-SMAC. Affinity for the BCR is an independent variable [34]. Antigens can combine TI-2 and TI-I properties. Bacterial flagellin is an example, flagellin is recognized by TLR5 [29}. Activated complement (C3d) present on many TI-2 antigens in vivo reacts with the complement receptor CD21 (on B or DC), which on B cells is linked to the activator CD19. In addition, the exact two-dimensional spacing of the repetitive B epitopes on a rigid TI-2 antigen most likely plays a role even in synapses, possibly by maximizing the effects of BCR-associated tyrosine kinases on the immunoreceptor tyrosine-based activation motifs (ITAM) of neighboring BCR [14].

T-dependent immune response

The T-dependent immune responses in the B follicles that lead to the formation of secondary follicles with germinal centers (GC), generate the repertoire of long-lived memory B cells [45, 84]. Like adaptive immunity in general, memory increases the probability of survival in a world in which repeated assaults by opportunistic pathogens are the rule. The GC is a sophisticated incubator for B cell expansion. For the typical generation of 103-104 progeny from one B cell, multiple rounds of proliferation induced by FDC-B and B - T synapses are required. Somatic V(D)J hypermuta-

(5)

tion and Ig isotype switching are particularly frequent during the GC reaction. Affin- ity maturation, which involves hypermutation together with cell selection, seems to exclusively occur in the GC under normal conditions.

Induction of the GC response

The T-dependent response requires "cognate" T - B interaction. For the initiation of the response, B cells must see, on a DC (or FDC), an antigen which, after internal- ization and enzymatic processing, generates a peptide (T epitope) presentable on MHC class II. Helper T cells must see antigen first on a DC, and a T cell must then recognize on a B cell the same peptide together with the same M H C class II that it has seen on a DC. The chemokine receptor CXCR5 participates in the homing of recirculating B cells to follicles; the ligand CXCL 13 is elaborated by FDC [13, 62]. Some studies in mice indicate that the B cell encounters the antigen on a DC for the first time when it travels from the high endothelial venules to the follicle through the T zone of the lymph node. The B cell then also encounters a specific helper T cell during its journey. Other data indicate that B cells first recognize the antigen on an FDC in the follicle [21]. FDC efficiently capture antigens coated with antibody (natural antibody or cross-reactive memory antibody) and/or complement C3d, which arrive through lymphatics. In this case, the B cell must encounter a T cell at the edge of (or within) the follicle. Both scenarios may occur depending on the site of the first encounter with antigen.

Depending on the nature of the antigen, a T-independent B response can be initi- ated in the GC [75]. Sooner or later a few activated helper T cells that also express CXCR5 - a species of T cell not fitting into the Thl/Th2 paradigm - migrate into the follicle [40, 62]. Initial costimulation of the T cell by the DC involves B7-1/2 interaction with CD28 constitutively expressed on the T cells. Activated T cells then express the inducible costimulator ICOS. The B cell has the ligand, B7RP-1. The B cell has to stimulate the T cell to elicit helper functions mediated by CD40L and other cell-bound or secreted T molecules [70]. The B cell also stimulates the T cell with OX40L; many TNF-family molecules (CD40L, CD27L, CD30L, OX40L, 4-1BBL, TNF, FasL, BAFF) and their receptors are involved in T-B collaboration [38]. Even- tually, about two or three B cells per GC generate expanded clones. The cells proliferate as large centroblasts in the dark zone of the GC. After a few rapid divisions, one every 6-8 h, they become smaller centrocytes and localize to the FDC network in the apical light zone where the GC T cells also reside. Here reactivation takes place. Obviously the antibody response should start early after infection by a pathogen. Thus, B cells leave the GC after various numbers of proliferation cycles, some B cells stay in the GC for 2 weeks or longer.

Switching and hypermutation

Hypermutation and Ig class switching occur in the proliferating cells. Both processes are mechanistically transcription-coupled. Hypermutation involves DNA double- strand breaks during transcription of V regions, followed by repair by components of DNA, "mismatch repair", in conjunction with several error-prone, lesion-bypass DNA polymerases [52, 69]. Polymerase Xl is an A - T mutator in this process, as

(6)

found in patients with variant xeroderma pigmentosum with a deficiency of this polymerase [81]. The specific signals inducing hypermutation are not yet known; CD40L is not essential [79].

Class switching occurs as a result of recombination between certain DNA regions located upstream of each Ig heavy-chain constant region. The essential regions re- main to be discovered; the classical tandem-repeat la switch region is important but not essential for switching starting from IgM [39]. Cytokine signals select target regions for switching, by inducing chromatin modification and generation of germline transcripts from the targeted regions [68]. DNA recombination then occurs, linking VDJ to a new constant region located further downstream than the one initially expressed. Valious components of non-homologous DNA end-joining that also participate in the V(D)J recombination are utilized (DNA end-binding proteins, DNA- dependent kinase, but RAGI/2 are not involved in the DNA cleavage). In humans with a "non-leaky" CD40L defect mutation, virtually no switch occurs; this is the Xqinked hyper-lgM syndrome. Activation-induced cytidine deaminase (AIC) deficiency gives an autosomal hyper-IgM syndrome as well as lack of hypermutation [58]. AIC could be an RNA-editing enzyme. Thus, related proteins resulting fiom this RNA editing may participate in both processes. In humans with severe combined immunodeficiency with normal B cells (i.e. T cell deficiency), it is also almost exclusively IgM that is produced.

Various cytokines target switching to distinct IgG subclass patterns; switching to IgE is well known to be dependent on interleukin-4 (IL-4) or IL-13. In mice with a conditional (Cre/loxP-mediated) transforming growth factor ~ (TGFI3) receptor deletion, which allows testing of this receptor function in adult animals, virtually no IgA switch occurs [I1]. In naive or memory human B cells from peripheral blood, TGF~I strongly induces IgAl and IgA2 germline transcripts, but IgA switching did not occur in the presence of DC [4], oligomeric CD40L or mouse EL-4 thymoma cells [80] in the presence of many cytokines tested (C. Werner-Favre, F. Bovia, R. Zubler, unpublished). It is possible that IgA switching depends on MALT-type microenvironmental components [41].

Cell selection in the GC

In conjunction with cell selection, hypermutation leads to antibody affinity maturation. Hypermutation carries the risk of producing autoantibodies. The general princi- ples of the cell selection for affinity maturation are well known. GC B cells are in an apoptosis-sensitive state; e.g. the anti-apoptotic protein Bcl-2 is down-regulated. The B cells need survival signals from antigen, FDC and T cells. The B cells that are crowded in the GC compete for antigen. Conditions leading to reduced apoptosis re- lax the cell selection and affinity maturation; they also increase autoantibody generation and the risk of follicular lymphomagenesis.

FasL/Fas-mediated apoptosis is important. Activated T and B cells express FasL and Fas. Because CD40L is a potent inducer of Fas, autoimmunity can even be found in CD40L deficiency [23]. FLIP protein competes with caspase-8 for the formation of the death-inducing signaling complex at the Fas death receptor. BCR cross-linking increases FLIP activity (in addition to various other anti-apoptotic effects, such as up-regulation of Bcl-xL). Murine B cells are protected from Fas death by FLIP hyperexpression [74]. Using lentiviral vectors [60], we found that human B cells trans-

(7)

duced with viral FLIPs are completely protected from the effect of oligomeric Fas ligand (E Bovia, R Salmon, C. Werner-Favre, et al., studies in progress). In contrast to B cells, DC express FLIP constitutively and are in fact activated by FasL [57]. B cell activation by FasL has not been found.

What normally happens when a BCR mutates to autoreactivity is less clear. If a B cell no longer presents those peptides that are recognized by the T cells in the GC, it should die out. T cells have a key control function. In mice expressing the influen- za virus hemagglutinin (HA) as self-antigen, in which naive anti-HA B cells are negatively selected, a strong anti-HA GC response and anti-HA B memory cell generation can be induced in the presence of virus-specific helper T cells [55]. B cells cross-reacting with self and the original antigen pose a problem. Possibly, HLA-DO, a modulator of peptide presentation on MHC II in B cells, is important in the control of autoimmunity [27].

Long-lived

p l a s m a cells and m e m o r y B cells

lg secretion does not occur in the GC; it would interfere with the cell selection for affinity maturation. Generally, proliferation signals inhibit differentiation in B cells. In proliferating B cells the transcription factor BSAP/Pax5 is a potent suppressor of other transcription factors required for the terminal differentiation of B cells into antibody-secreting plasma cells, such as XBP-I and Blimp-l [56, 73]. The gene repres- sor Bcl-6 acts on various genes for cell-cycle control [Bcl-6 can become oncogenic in follicular lymphoma by the V(D)J hypermutation mechanism] as well as genes for plasmocytic differentiation [65].

B cells that have been activated in an immune response (T-dependent or T-independent) permanently acquire CD27 expression. CD27 cross-linking in conjunction with [L-10 and IL-2 leads to B cell differentiation [2]. Since some T as well as B cells in the GC express CD27L, a certain proportion of centrocytes can leave the GC already committed for differentiation, equipped with homing molecules for the medullary chords of lymph nodes, the bone marrow, the mucosa and other sites to which the plasma cell precursors disseminate. Since recirculating CD27 + B cells can also be readily induced in vitro to become plasma cells by BCR cross-linking with SAC [44] and cytokines, such as IL-2 and IL-10, which they produce themselves [32, 84 ], other post-GC B cells could undergo plasmocytic differentiation in various tissues in the presence of antigen. Syndecan-1 is a marker for plasma cells.

Some plasma cells have a long life span [45, 84]. In humans, several months of systemic treatment with anti-B-cell (anti-CD20) antibody for lymphoma does not significantly reduce the serum Ig levels [42]. Other plasma cells are short-lived (days to weeks), in particular those that appear first during the T-dependent response and home to the medullary chords in the local lymph node. The life span could depend on the microenvironment. It is also possible that repeated restimulation in synapses favors longevity of B cells, as has been proposed for T cells [36]. This would explain why the GC reaction generates plasma and memory cells with a long life span.

As with memory T cells, persisting antigen is not required for survival of memory B cells [43]. Recirculating human memory B cells maintain a high expression of Bcl-xL anti-apoptotic protein, in contrast to naive B cells [8]. When such cells are put in culture without stimulation, the Bcl-xL mRNA decreases rapidly, indicating that a high level is maintained by exogenous signals. For memory T cells in mice,

(8)

IL-15, but not IL-2, provides an important survival signal [35]. Memory cells (B or T) undergo antigen-independent cell division from time to time; they show a capacity for self-renewal and homeostatic proliferation.

S w i t c h e d and n o n - s w i t c h e d C D 2 7 § B cells

The human recirculating CD27+ B cells, but not the naive B cells induced to express CD27 by SAC in culture, have acquired the capacity for plasmocytic differentiation in assays in vitro [44, 80]. This reveals some functional maturity, which may require T - B or DC-B synapse formation and/or still unknown B cell activators. About 45% of recirculating B ceils in adults are CD27 +. Among these ceils, Ig-switched cells, IgM + cells that lack IgD (IgM-only cells), and IgM+IgD § cells occur in similar proportions, and more than 90% of all these cells can'y V(D)J mutations [33]. Thus, a high proportion of non-switched, mutated B cells occur in humans.

In patients with complete CD40L functional deficiency, both the switched and the IgM-only cells are lacking, whereas some V(D)J mutated IgM+IgD § CD27 + cells are produced, although in most cases at low levels (1%-2% of peripheral blood B cells in six of eight patients, 4% and, for unknown reasons, 60% in the two others) compared to normal controls (7%-10%) [79]. Follicles are severely disorganized in CD40L deficiency; CD40L is also required to maintain FDC. Thus, some hypermutation could either still occur in such tissue or take place elsewhere, i.e. in the splenic marginal zone [79]. In some species, like sheep, diversification by V(D)J mutation occurs during B cell ontogeny, mainly in the GALT. The marginal zones in spleen, Peyer's patches and lymph nodes are indeed an important reservoir of V(D)J-mutated B cells in adults [15, 7t, 72]. However, in hyper-IgM syndrome one expects a compensatory increase of the T-independent marginal zone B celt responses to infec- tious agents. Thus, the low levels of recirculating mutated B cells in this condition indicate that these normally are mostly GC-derived. Therefore, in normal individu- als, when they are not just undergoing an acute immune response (T-dependent or T-independent) with high blood levels of rapidly tissue-homing CD27 + plasma cell precursors, the recireulating CD27 § cells are essentially long-lived post-GC memory B cells. The high proportion of memory B cells may be due to the longer life span of humans compared to mice.

Ig s w i t c h c a p a c i t y of CD27 § B cell s u b s e t s

Recently we investigated the Ig switch capacity of the three CD27 + B subsets [80]. IgM+IgD § cells switch to all IgG subclasses to the same extent as naive ceils. In contrast, IgM-only cells have a very low switch activity. This suggests a switch defect, such as could occur when attempted switching leads to loss of certain p DNA switch regions. Lack of this subset in CD40L deficiency [79] accords with some switch- related mechanism. The CD27 + cells that had already switched to IgG in vivo showed no increase of IgG4 in response to IL-4. Thus, although it is known that secondary switching to further-downstream constant regions can occur [68], at least some switch options are strongly reduced in cells that have already switched.

The generation of these three post-GC subsets may reflect evolutionary pressure for optimization of immune responses. A tendency for "isotype stabilization" in

(9)

switched cells would ensure that a secondary response to the same pathogen generates those isotypes which have been instructed by signals of natural immunity during the primary response [82]. The GC also produces V(D)J-mutated B cells that can cross-react with novel pathogens. Fully conserved switch capacity is an advantage when the cells respond to novel pathogens.

Moreover, affinity maturation of IgM antibodies seems to be very important, although the IgM are pentameric antibodies. Many IgM antibodies to the polysaccharides of encapsulated bacteria in humans carry V(D)J mutations and, at least for some of these antigens, depending on the germline V repertoire, mutations increase the antibody affinity [1]. A cellular receptor for polymeric IgA and IgM (Fccx/laR) mediating phagocytosis of antibody-coated bacteria has recently been found [66], ex- tending the effector functions of IgM beyond antigen neutralization and complement activation. Also, deficiency of serum IgM in mice predisposes them to development of IgG autoantibodies; maintenance of a certain IgM/IgG ratio seems to be important [17]. IgM has a short half-life in vivo but, compared to monomeric IgG, it takes five times more plasma cells to produce an equal number of IgM molecules. Therefore, the immune system might utilize active switch suppression in the GC, and not just rely on the probabilistic nature of switching. GC T cells expressing CD30L can suppress switching in human B cells [12]. Interestingly, this preferentially affects non- antigen-selected cells, i.e. possibly B cells that have mutated and no longer recognize the GC-response-inducing antigen, but which can serve as functionally experienced cells reacting with novel pathogens. Loss of switching capacity in the IgM-only cells may also be a means of isotype stabilization to maintain a certain IgM level.

M e m o r y B cells for T-independent responses

The capacity of the immune system to generate T-independent B responses is important in the case of antigens that can not induce a classical T - B collaboration, such as polysaccharides or phospholipids. Such responses can be enhanced by pathogen receptors of natural immunity. Potentially, u cells recognizing non-peptide antigens on non-classical MHC molecules can also participate [76]. However, it is important to realize that not only naive but also the functionally experienced memory B cells can participate in T-independent responses (Fig. 1). As mentioned above, compared to naive cells, human memory B cells show a much enhanced response to T-independent stimulation. The memory B cell pool is quantitatively important in adults. Depending on the pathogen and the primary BCR specificity repertoire, it is more the functional maturity, the novel cross-reactivities or the affinity maturation of the memory B pool that leads to enhanced T-independent responses. Even pathogens that induce an almost exclusively T-independent primary response can, after repeated exposure, cause an accumulation of specific memory B cells. This occurs when a small proportion of the T-independent antigen forms covalent or non-covalent asso- ciations with proteins that allow for a T-dependent GC response with pathogen- specific or cross-reactive T helper cells. This is natural immunization corresponding to a low-dose polysaccharide-protein conjugate vaccine. The clinically relevant im- maturity of the T-independent response to the polysaccharides of encapsulated bacteria in infants aged less than 2 years [1] most likely reflects the time required to generate memory B cells. Progressively, the marginal zones in spleen and lymph nodes become enriched in post-GC cells [ 15, 71, 72].

(10)

Adaptive humoral immune responses

Natural antitxx:lies T-independent antibodies T-dependent an~boOies

Pnmary Memory Memory Primary

@

@ |

@ @

B cells for marginal zones (MZ) Clonal deletion ~ Receptor editing B ceils for fo(lic!es

Pnrnary V(D)J rearrangements with non-random V region ~[izalion

Fig. 1. Participation of primary and memory B cell subsets in T-dependent and T-independent humoral immune response. Bn naive B cells, Bm memory B cells; PL plasma cells, DC dendritic cells; F D C follicular dendritic cells; Ttl helper T cells; M Z marginal zones; GC germinal centers

A cytokine receptor for T-independent B responses

The B lymphocyte activator of the TNF family (BAFF), also called B lymphocyte stimulator (BEyS), THANK, TALL-1 or zTNF4, has recently been discovered in humans and mice [47, 63, 67, 77]. This cytokine acts on two known receptors, BCMA, which seems to be B-specific, and TACI found on B and T cells [67]. One function of BAFF is to act as a survival factor for immature B cells emigrating from the bone marrow to the spleen, i.e. for the cells in which self-tolerance has to be established [6]. BAFF hyperexpression in transgenic mice causes B cell hyperplasia and autoimmunity. TACI-Ig fusion protein, used as a decoy receptor, decreases autoantibody production and autoimmune disease in NZBWFI mice [25]. A P R I L a related TNF- family factor, is produced by various tumor cells and acts on both BAFF receptors, revealing new links between cancer and autoimmunity [77]. Antibodies to APRIL have also been found to reduce T cell activation in mice [67].

Regarding the T-independent B cell response, BAFF is secreted by interferon-~/- activated monocytes and DC [47]. TACI-knockout mice have a severe defect of TI-2 B responses [9]. This is the first example of an important cytokine receptor for the T- independent response. Surprisingly, TACI-knockout mice also have twice as many B cells as control mice [9]. The different BAFF receptors may have opposite regulatory functions during B cell development.

(11)

Conclusions

Continuous production of new B ceils maintains primary BCR specificity repertoires. B l B cells are positively selected by weak self-reactivity. Their differentiation into plasma cells producing natural antibodies of mainly IgM and IgA class in mice occurs in the MALT and spleen, with participation of y/iS-T cells, Conventional naive B cell subsets home to either the marginal zones or the B follicles of secondary lymphoid organs, the preferential microenvironments for T-independent and T-dependent primary B responses respectively. The T-dependent responses in the follicular germinal centers progressively build up the repertoire of long-lived, functionally experienced and V(D)J-mutated memory B cells, which eventually represent about half of all B cells in peripheral blood and the marginal zones in adult humans. These cells can undergo T-dependent and T-independent secondary responses as well as cross-react with novel pathogens. They express CD27 and comprise three subsets in similar proportions: (1) the Ig-class-switched cells, which exhibit a low second- ,'u-y-switching capacity to at least some isotypes and thus conserve isotype patterns for secondary responses, (2) the IgM-onty cells, which have lost switching capacity and assure high-affinity IgM production for complement- and Fcct/~aR-mediated de- fenses, and (3) the IgM+IgD + ceils, which conserve full switching capacity and may therefore be optimal for cross-reactive responses to novel pathogens, Because of their functional maturity and long life span, the post-GC memory B cells are important for the T-independent B responses in humans.

Acknowledgement. The author's original work reported in this review was supported by a grant from the Swiss National Foundation for Scientific Research

References

1, Adderson EE (2001) Antibody repertoires in infants and adults: effects ofT-independent and T-dependent immunizations. Semin lmrnunopathol 23

2. Agematsu K, Nagumo H, Oguchi Y, Nakazawa T, Fukushima K, Yasui K, "[to S, Kobata T, Morimoto C, Komiyama A 0998) Generation of plasma ceils from peripheral blood memory B cells: synergistic effect of interleukin-10 and CD27/CD70 interaction. Blood 91:173

3. Ansel KM, Ngo VN, Hyman PL, Luther SA, Fisrster R, Sedgwick JD, Browning JL, Lipp M, Cyster JG (2000) A chemokine-driven positive feedback loop organizes lymphoid follicles. Nature 406:309 4. Arrighi J-F, tlauser C, Chapuis BI Zubler RH, Kindler V (1999) Long-term culture of human CD34 §

progenitors with FLT3-ligand, thrombopoietin, and stem cell factor induces extensive amplification of a CD34-CDI4- and a CD34-CD 14 § dendritic cell precursor. Blood 93:2244

5. Batista FD, Iber D, Neuberger MS (2001) B cells acquire antigen from target cells after synapse formation. Nature 411:489

6. Batten M, Groom J, Cachero TG, Qian F, Schneider E Tschopp J, Browning JL, Mackay F (2000) BAFF mediates survival of peripheral immature B lymphocytes. J Exp Med 192:1453

7. Bikah G, Carey J, Ciallella JR, Tarakhovsky A, Bondada S (1996) CD5-mediated negative regulation of antigen receptor-induced growth signals in B-t B cells. Science 274:1906

8. Bovia E Nabili-Tehrani AC, Werner-Favre C, Barnet M, Kindler V, Zubler RH (1998) Quiescent memory B cells in hurnan peripheral blood co-express bcl-2 and bcl-x L anti-apoptotic proteins at high levels. Eur J lmmunol 28:4418

9. Btitow G-U yon, Deursen JM van, Brain RJ (2001) Regulation of the T-independent humoral response by TACI. Immunity 14:573

10. Casellas R, Shih T-A Y, Kleinewietfeld M, Rakonjac J, Nemazee D, Rajewsky K, Nussenzweig MC (2001) Contribution of receptor editing to the antibody repertoire. Science 291 : 1541

(12)

11. Cazac BB, Roes J (2000) TGF-13 receptor controls B cell responsiveness and induction of IgA in vivo. Immunity 13:443

12. Cerutti A, Schaffer A, Shah S, Zan H, Liou H-C, Goodwin RG, Casali P (1998) CD30 is a CD40- inducible molecule that negatively regulates CD40-mediated immunoglobulin class switching in non- antigen-selected human B cells. Immunity 9:247

13. Cyster JG, Ansel KM, Reif K, Ekland EH, Hyman PL, Tang HL, Luther SA, Ngo VN (2000) Follicu- lar stromal ceils and lymphocyte homing to follicles. Immunol Rev 176:181

14. DeFranco AL (2000) B-cell activation 2000. lmmunol Rev 176:5

15 Dunn-Walters DK, lsaacson PG, Spencer J (t995) Analysis of mutations in immunoglobulin heavy chain variable region genes of microdissected marginal zone (MGZ) B cells suggests that the MGZ of human spleen is a reservoir of memory B cells. J Exp Med 182:559

16. Dustin ML, Chan AC (2000) Signaling takes shape in the immune system. Cell 103:283

17. Ehrenstein MR, Cook HI', Neuberger (2000) Deficiency in serum immunoglobulin ([g)M predisposes to development of lgG autoantibodies. ] Exp Med 191:1253

18. Fagarasan S, Honjo T (2000) T-independent immune response: new aspects of B cell biology. Science 290:89

19. FIo TH, Halaas O, qbrp S, Ryan L, Lien E, Dybdahl B, Sundan A, Espevik T (2001) Differential expression of Toll-like receptor 2 in human cells. I Leukoc Biol 69:474

20. Fruman DA, Satterthwaite AB, Witte ON (2000) Xid-like phenotypes: a B cell signalosome takes shape. Immunity 13:1

21. Garside P, Ingulli E, Merica RR, .l'ohason JG, Noelle RJ, Jenkins MK (1998) Visualization of specific B and T lymphocyte interactions in the lymph node. Science 281:96

22. Golovkina TV, Shlomchik M, Hannum L, Chervonsky A (1999) Organogenic role of B tymphocytes in mucosal immunity. Science 286: t 965

23. Goodnow CC (2001) Pathways for self-tolerance and the treatment of autoimmune diseases. Lancet 357:2115

24. Grimaitre M, Werner-Favre C, Kindler V, Zubler Rtt (1997) Human naive B cells cultured with EL-4 T cells mimic a germinal center-related B cell stage before generating plasma cells. Concordant changes in Bcl-2 protein and messenger RNA levels. Eur J Immunol 27:199

25. Gross JA, Johnston J, Mudri S, Enselman R, Dillon SR, Madden K, Xu W, Parrish-Novak J, Foster D, Lofton-Day C. Moore M, Littau A, Grossman A, Haugen H, Foley K, Blumberg H, Harrison K, Kindsvogel W. Clegg CH (2000) TACI and BCMA are receptors for a TNF homologue implicated in B-cell autoimmune disease. Nature 404:995

26. Guinamard R, Okigaki M, Schlessinger J, Ravetch JV (2000) Absence of marginal zone B cells in Pyk-2-deficient mice defines their role in the humoral response. Nat Immunol 1:9

27. tfam M van, Lith M van, Lillemeier B, Tjin E, Griineberg U, Rahman D, Pastoors I,, Meijgaarden K van, Roucard C. Trowsdale J, Ottenhoff T, Pappin D, Neefjes J (2000) Modulation of the major histocompatibility complex class II-associated peptide repertoire by human histocompatibility leukocyte antigen (HLA)-DO. J Exp Med 191:1127

28. Hayakawa K, Asano M, Shinton SA, Gui M, Allman D, Stewart CL, Silver J, Hardy RR (1999) Posi- tive selection of natural autoreactive B cells. Science 285:113

29. Hayashi E Smith KD, Ozinsky A, Hawn TR, Yi EC, Goodlett DR, Eng JK, Akira S, Underhill DM, Aderem A (2001) The innate immune response to bacterial flageIlin is mediated by Toll-like receptor 5. Nature 410:1099

30. Hippen KL, Tze LE, Behrens TW (2000) CD5 maintains tolerance in anergic B ceils. J Exp Med 191:883

31. Kindler V, Zubler RH (1997) Memory, but not naive, peripheral blood 13 lymphocytes differentiate into Ig-secreting cells after CD40 ligation and costimulation with IL-4 and the differentiation factors IL-2, IL-10, and IL-3_ J Immunol 159:2085

32. Kindler V, Matthes T, Jeannin P, Zubler RH (1995) Interleukin-2 secretion by human B lymphocytes occurs as a late event and requires additional stimulation after CD40 cross linking. Eur J Immnnol 25:1239

33. Klein U, Rajewsky K, Ki.ippers R (1998) Human immunoglobulin (Ig)M+IgD + peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: CD27 as a general marker for somatically mutated (memory) B cells. J Exp Med 188:1679

34. Kouskoff V, Famiglietti S, Lacaud G, Lang P, Rider JE, Kay BK, Cambier JC, Nemazee D (1998) Antigens varying in affinity for the B cell receptor induce differential B lymphocyte responses. J Exp Med 188:1453

(13)

35. Ku CC, Murakami M, Sakamoto A, Kappler J, Marrack P (2000) Control of homeostasis of CD8 + memory T cells by opposing cytokines. Science 288:675

36. Lanzavecchia A, Sallusto F (2000) Dynamics o f t lymphocyte responses: intermediates, effectors, and memory cells. Science 290:92

37. Li BL, Zhang X, Kovacic S, Long AJ, Bourque K, Wood CR, Choi YS (2000) Identification of a human follicular dendftic cell molecule that stimulates germinal center B cell growth. J Exp Med 191:1077

38. Locksley RM, Killeen N, Lenardo MJ (2001) The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell 104:487

39. Luby TM, Schrader CE, Stavnezer J, Selsing E (2001) The la switch region tandem repeats are important, but not required, for antibody class switch recombination. J Exp/Vled 193:159

40. Mackay CR (2000) Follicular homing T helper (Th) cells and the Thl/Th2 paradigm (Commentary). J Exp Med 192:F31

4l. Macpherson AJS. Lamarre A, McCoy K, Harriman GR, Odermatt B, Dougan G, Hengartner H, Zinkernagel RM (2001) lgA production without la or ,5 chain expression in developing B cells. Nat lmmunol 2:625

42. Maloney DG, Grillo-Lopez A J, White CA, Bodkin D, Schilder RJ, Neidhart JA, Janakiraman N, Foon KA, Kdes TM, Dallaire BK, Wey K, Royston I, Davis T, Levy R (1997) IDEC-C2B8 (Rituximab) anti-CD20 monoclonal antibody therapy in patients with relapsed low-grade non-Hodgkin's lymphoma. Blood 90:2188

43. Maruyama M, Lain K-E Rajewsky K (2000) Memory B-ceil persistence is independent of persisting immunizing antigen~ Nature 407:636

44. Maurer D. Fischer GF, Fae I, Majdic O, Stuhimeier K, Jeney N yon, Hotter W, Knapp W (1992) IgM and [gG but not cytokine secretion is restricted to the CD27+ B lymphocyte subset. J Immunol

t48:3700

45. McHeyzer-Witliams kJ, Driver DJ, McI-teyzer-Wiltiams MG (2001) Germinal center reaction. Curr Opin Hematol 8:52

46. Monks CRE Freiberg BA, Kupfer H, Sciaky N, Kupfer A (1998) Three-dimensional segregation of supramolecular activation clusters ira T cells. Nature 395:82

47. Moore PA, Belvedere O, Orr A, Pieri K, LaFleur DW, Feng P, Soppet D, Charters M, Gentz R, Parmelee D, Li Y, Galperina O, Girl J, Roscbke V, Nardelli B, Carrell J, Sosnovtseva S, Greenfield W, Ruben SM, Olsen HS, Fikes J, Hilbert DM (1999) Blys: Member of the rumor necrosis factor family and B lymphocyte stimulator. Science 285:260

48. Morse HC 3rd, Kearney JF, Isaacson PSG, Carroll M, Fredrickson TN, Jaffe ES (2001.) Cells of the marginal zone - origins, function and neoplasia. Leuk Res 25:169

49. Mazio M, Bosisio D, Polentarutti N, D'Amico G, Stoppacciaro A, Mancinelli R, van't Veer C, Penton-Rol G, Ruco LP, Allavena P, Mantovani A (2000) Differential expression and regulation of toll-like receptors (TLR) in human leukocytes: selective expression of TLR3 in dendritic cells. J ImmunoI 164:5998

50. Ochsenbein AE Fehr T, Lutz C, Surer M, Brombacher F, Hengartener H, Zinkernagel RM (1999) Control of early viral and bacterial distribution and disease by natural antibodies. Science 286:2156 51. Oldenborg PA, Zheleznyak A, Fang YF, Lagenaur CF, Gresham HD, Lindberg FP (2000) Role of

CD47 as a marker of self on red blood cells. Science 288:2051

52. Poharatsky V. Goodman MF, Scharff MD (2000) Error-prone candidates vie for somatic mutation. J Exp Med 192:F27

53. Pulendran B, Palucka K, Banchereau J (2001) Sensing pathogens and tuning immune responses. Science 293:253

54. Ravetch JV, Lanier LL (2000) Immune inhibitory receptors. Science 290:84

55. Reed A J, Riley ME Caton AJ (2000) Virus-induced maturation and activation of autoreactive memory B cells. J Exp Med 192:1763

56. Reimold AM, Iwakoshi NN, Manis J, Vallabhajosyula R Szomolanyi-Tsuda E, Gravallese EM, Friend D, Grusby MJ, Alt F, Glimcher LH (2001.) Plasma cell differentiation requires the transcription factor XBP-I. Nature 412:300

57. Rescigno M, Piguet V, Valzasina B, Lens S, Zubler R, French L, Kindler K, Tschopp J, Ricciardi- Castagnoli P (2000) Fas engagement induces the maturation of dendritic cells (DCs), the release of interleukin (IL)-I[3, and the production of interferon "r in the absence of IL-12 during DC-T cell cognate interaction: a new role for Fas ligand in inflammatory responses. J Exp Med 192:1661

(14)

58 Revy E Muto T. Levy Y, Geissmann F, Plebani A, Sanal O, Catalan N, Forveille M, Dufourcq- Labelouse R, Gennery A, Tezcan I, Ersoy F, Kayserili H, Ugazio AG. Brousse N, Muramatsu M. Notarangelo LD, Kinoshita K, Honjo T, Fischer A. Durandy A (2000) Activation-induced cytidine deaminase (AID) deficiency causes the antosomal recessive form of the hyper-IgM syndrome (HIGM2). Cell 102:54l

59. Richards SJ, Morgan GJ, Hillmen P (2000) lmmunophenotypic analysis of B cells in PNH: insights into the generation of circulating naive and memory B cells. Blood 96:3522

60. Salmon E Kindler V, Ducrey O; Chapuis B, Zubler RH, Trono D (2000) High-level transgene expression in human hematopoietic progenitors and differentiated blood lineages after transduction with im- proved lentiviral vectors. Blood 96:3392

61. Sandel PC, Monroe JG (1999) Negative selection of immature B cells by receptor editing or deletion is determined by site of antigen encounter. Immunity 10:289

62. Schaerli P, Willimann K, Lang AB, Lipp M, Loetscher E Moser B (2000) CXC chemokine receptor 5 expression defines follicular homing T cells with B cell helper function. J Exp Med 192:1553 63. Schneider P, MacKay E Sterner V, Hofmann K, Bodmer J-L, Holler N, Ambrose C, Lawton P, Bixler

S, Acha-Orbea H, Valmori D, Romero P, Werner-Favre C, Zubler RH, Browning JL, Tschopp J (1999) BAFE a novel ligand of the tumor necrosis factor family, stimulates B cell growth. J Exp Med

189:1747

64. Seddon B, Legname G, Tomlinson P, Zamoyska R (2000) Long-term survival but impaired homeostatic proliferation of na T cells in the absence of p56 lck. Science 290:127

65. Shaffer AL, Yu X, He Y, Boldrick J, Chart EE Staudt LM (2000) BCL-6 represses genes that function in lymphocyte differentiation, inflammation, and cell cycle control. Immunity 13:199

66. Shibnya A, Sakamoto N, Shimizu Y, Shibnya K, Osawa M, Hiroyama t-, Eyre HJ, Suthertand GR, Endo Y, Fujita T, Miyabayashi T, Sakano S, Tsuji ~l; Nakayama E, Phillips JH, Lanier LL, Nakauchi H (2000) Fc alpha/mu receptor mediates endocytosis of IgM-coated microbes. Nat lmrnunol 1:441 67. Siegel RM, Lenardo MJ (2001) To B or not to B: TNF family signaling in lymphocytes. Nat Immunot

2:577

68. Snapper CM, Marcu KB, Zelazowski P (1997) The immunoglobulin class switch: Beyond "Accessi- bility". Immunity 6:217

69. Storb U (2001) DNA polymerases in immunity: profiting from errors. Nat lmmunol 2:484

70. Tafuri A, Shahinian A, Bladt F, Yoshinaga SK, Jordana M, Wakeham A, Boucher L-M, Bouchard D, Chan VSP, Duncan G, Odermatt B, Ho A, hie A, Horan T, Whoriskey JS, Pawson T, Penninger JM, Ohashi PS, Mak TW (2001) ICOS is essential for effective T-helper-cell responses. Nature 409:105 71. Tangye SG, Liu Y-J, Aversa G, Phillips JH, Vries JE de (1998) Identification of functional human

splenic memory B cells by expression of CD148 and CD27. J Exp Med 188:1691

72. Tierens A, Delabie J, Michiels L, Vandenberghe R De Wolf-Peeters C (1999) Marginal-zone B cells in the human lymph node and spleen show somatic hypermutations and display clonal expansion. Blood 93:226

73. Usui T, Wakatsuki Y, Matsunaga Y, Kenko S, Kosek H, Kita T (1997) Overexpression of B cell- specific activator protein (BSAP/Pax-5) in a late B cell is sufficient to suppress differentiation to an Ig high producer cell with plasma cell phenotype. J lmmunol 158:3197

74. Van Parijs L, Rel-aeli Y, Abbas AK, Baltimore D (1999) Autoimmunity as a consequence of retrovirns- mediated expression of C-FLIP in lymphocytes. Immunity 11:763

75. Vinuesa CG de. Cook MC, Ball J, Drew M, Sunners Y, Cascalho M, Wabl M, Klau GG, MacLennan 1C (2000) Germinal centers without T cells. J Exp Med 191 : 485

76. Vos Q, Lees A, Wu Z-Q, Snapper CM, Mond JJ (2000) B-cell activation by T-cell-independent type 2 antigens as an integral part of the humoral immune response to pathogenic microorganisms. Immunol Rev 176:154

77. Ware CF (2000) APRIL and BAFF connect autoimmunity and cancer (Comment~y). J Exp Med 192:F35

78. Watanabe N, Ikuta K, Fagarasan S, Yazumi S, Chiba T, Honjo T (2000) Migration and differentiation of autoreactive B-l cells induced by activated "f/6 T cells in antierythrocyte immunoglobulin transgenic mice. I Ex p Med 192:t577

79. Wcller S, Faili A, Garcia C, Braun MC, Le Deist F. Saint Basile G de, Hermine O, Fischer A, Reynaud C-A, Weill J-C (2001) CD40-CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans. Proc Nat[ Acad Sci USA 98:1166

(15)

80. Werner-Favre C, Bovia F, Schneider P, Holler N, Barnet M, Kindler V, Tschopp J, Zub]er RH (2001) IgG subclass switch capacity is low in switched and in IgM-only, but high in IgD+IgM § post-germinal center (CD27 § human B ceils. Eur J Immunot 31:243

81. Zeng X, Winter DB, Kasmer C, Kraemer KH, Lehmann AR, Gearhart PJ (2001) DNA polymerase r I is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat lmmunol 2:537 82. Zhang K, Cbeah HK, Saxon A (1995) Secondary deletional recombination of rearranged switch region

in lg isotype-switched B cells. A mechanism for isotype stabilization. J lmmunol 154:2237

83. Zinkernagel RM, Hengarmer H (2001) Regulation of the immune response by antigen. Science 293:251

84. Zubler RH (1997) Key differentiation steps in normal B cells and in myeloma ceils. Semin Hematol 34 [Snppl l]:13